New process fills technology gap in removing [H.sub.2]S from gas – hydrogen sulfide, natural gas – Statistical Data Included
Kenneth E. McIntush
Nonaqueous sulfur recovery treats high-pressure gas without separate upstream [H.sub.2]S process, making production with medium sulfur volumes practical to market
Described here are the three existing aqueous-type processes for treating [H.sub.2]S in higher-pressure natural gas streams, and a new, nonaqueous, proprietary sulfur recovery process called CrystaSulf, marketed by CrystaTech, Inc., of Austin, Texas. The four processes are overviewed, along with their applications in three “size” ranges: small, large and the newer, more challenging medium. Benefits of the new process are outlined, along with cost comparisons for treating high-pressure sour gas with three different processes. Conclusions note that initial design work has begun on the first of two commercial CrystaSulf units.
INTRODUCTION
As much as 15-25% of the natural gas in the U.S. may contain hydrogen sulfide ([H.sub.2]S) [1,2] Worldwide, the percentage could be as high as 30% [3] Pipeline specs limit [H.sub.2]S concentrations to low levels (typically 4 ppm in the U.S.). The need for more cost-effective approaches to remove [H.sub.2]S is evident, as new drilling focuses on deep gas formations that tend to be “sour.” [4.5]
Carbon dioxide ([CO.sub.2]) is often found in conjunction with [H.sub.2]S However, most U.S. gas that contains [H.sub.2]S has [CO.sub.2] concentrations below 3%, the common U.S. [CO.sub.2] pipeline spec. [1] Even higher concentrations may be allowed in other countries. Although the [CO.sub.2] may not need to be removed, its mere presence has historically complicated [H.sub.2]S removal, because [CO.sub.2] adversely affects most [H.sub.2]S removal and sulfur recovery processes.
THREE SIZE RANGES
Prior to comprehensive regulation of [SO.sub.2] and [H.sub.2]S emissions, the market for processes to remove [H.sub.2]S and/or recover sulfur from gas streams could have been divided roughly into two primary niches, small- and large-size applications. The small-size niche-less than 0.1 long tons sulfur per day (LTPD)–was served with nonregenerable scavenging chemicals. [6] Large-sized applications were served by combinations of amine and modified Claus processes. Any applications that fell in between frequently employed [H.sub.2]S removal with an amine system, followed by venting or flaring the acid gases.
After [SO.sub.2] and [H.sub.2]S emission limits became stricter, high sulfur removal and recovery efficiencies became a requirement. The option of venting or flaring [H.sub.2]S was greatly limited. For applications outside the small-size range, using amine/Claus combinations began to involve additional process steps and higher costs to meet the new sulfur recovery requirements. As a result, a medium-size sulfur recovery niche formed for sulfur throughputs from about 0.1 to 30 LTPD.
Treating natural gas in this mediumsize niche has proved challenging, especially for applications at elevated pressures, i.e., anything over roughly 100 psig. Aqueous-iron liquid redox processes appeared to be the economic choice for treating this gas. However, it is well documented that these plants have significant operability and reliability problems, and they have been applied primarily at low pressures (less than 100 [psig). [7,8,9]
No liquid redox technology has been demonstrated to reliably treat gas streams where the gas pressure is greater than around 300 psig, unless special and generally uneconomic design approaches were used. Caustic processes with bacterial regeneration have also been proposed for treating high-pressure gas, but they are unproven commercially on high-pressure gas, and they are likely to have drawbacks similar to the aqueous liquid redox processes. When considering high-pressure treating, it is important to be aware that aqueous sulfur recovery systems have inherent limitations, especially at high [CO.sub.2] concentrations and pressures.
As a result of the problems encountered with the aqueous sulfur recovery systems, most processes that appeared to target this niche have made little market penetration. The lack of acceptable alternatives means that gas producers often use expensive amine/Claus/tail gas treating (TGT) combinations or do not produce the gas.
FOUR POTENTIAL PROCESSES
Gas transmission lines frequently operate at 600 psig or higher, so treating at these pressures is usually a requirement. Some of the primary features that make a process suitable for high-pressure natural gas operation are:
* Ability to remove [H.sub.2]S to 4 ppm
* Unaffected by high [CO.sub.2]
* Low liquid circulation rates (pumping costs)
* Good HP pump performance
* Low foaming tendency
* Low plugging tendency, and
* Low corrosivity.
Numerous processes and process combinations have been proposed for treating high-pressure gas including these four:
1. Aqueous amine/Claus/TGT combinations
2. Aqueous sulfur recovery processes, including: aqueous-iron liquid redox; and aqueous caustic with bacterial regeneration
3. Amine/aqueous-iron liquid redox combinations, and
4. The nonaqueous CrystaSulf process.
This article discusses each of these processes and how it performs relative to the desirable process features listed above.
Aqueous amine/Claus/TGT combinations. Fig. 1 shows a simplified flow diagram for an amine/Claus/TGT combination. In this combination, the amine-unit absorber is the only high-pressure vessel. Essentially, all the [H.sub.2]S and some or all of the [CO.sub.2] are absorbed into an aqueous alkanolamine (amine) solution. The [H.sub.2]S and [CO.sub.2] come out of the amine system in the low-pressure amine regenerator column and are then fed to the Claus unit, which converts most of the [H.sub.2]S to molten elemental sulfur. Tail gas from the Claus unit is then fed to a TGT unit to clean up any remaining sulfur species. This combination of processes is generally the most economic choice for applications with more than about 30 LTPD of sulfur.
The amine/Claus/TGT combination has most of the features desirable for high-pressure natural gas treatment, and it is proven reliable. However, when considering medium-sized applications, capital costs are high, relative to some of the other proposed options. Further, costs of the amine/Claus/TGT combination rise even more when high [CO.sub.2] to [H.sub.2]S ratios are encountered. Therefore, it is often uneconomical to produce gas containing medium amounts of sulfur with this approach.
Aqueous sulfur recovery processes. Several types of aqueous sulfur recovery systems have been proposed for use on high-pressure natural gas. Most were originally developed for low-pressure applications and, later, attempts were made to adapt them to high-pressure applications. This class of processes includes liquid redox systems and caustic processes with bacterial regeneration, among others. Systems originally developed for low-pressure applications, e.g., Stretford and numerous versions of chelated iron processes, have generally not performed well at elevated pressures. These systems have some inherent flaws for high-pressure operation, especially in the presence of significant [CO.sub.2] partial pressures.
This section first discusses some considerations generally applicable to all of the aqueous sulfur recovery systems. Then, two specific types are discussed: aqueous-iron liquid redox arid aqueous caustic with bacterial regeneration.
Effect of [CO.sub.2]. Fig. 2 shows conditions of [CO.sub.2] partial pressure and pH under which sodium bicarbonate precipitation occurs, for an ideal system of water with sodium ions in equilibrium with [CO.sub.2] at various [CO.sub.2] partial pressures (calculated for a temperature of 40[degrees]C using the ACAPP electrolyte program [10]). The general effect of pH on [H.sub.2]S removal for one aqueous sulfur recovery application is also shown.
As Fig. 2 shows, if liquid in equilibrium with the [CO.sub.2] (i.e., liquid in the absorber) has a pH of 8, sodium bicarbonate solids precipitate, unless the [CO.sub.2] partial pressure is 0.5 atm or lower. For pH 9, sodium bicarbonate precipitates, unless [CO.sub.2] partial pressure is 0.1 atm or lower. Precipitation of solids in the absorber is generally undesirable and should be avoided to prevent absorber fouling, formation of particle-stabilized foams, etc.
In general, 1) inability to run at high pH without possible precipitation problems, and 2) poor removal at lower pH, greatly limit the applicability of caustic scrubbing systems for high-pressure gas treating when [CO.sub.2] partial pressures are high. Dilute aqueous-iron systems may also see sodium bicarbonate precipitation, if NaOH is used for pH control. In one example, a dilute chelated iron system applied to a gas stream with about 0.5 atm [CO.sub.2] partial pressure was experiencing poor [H.sub.2]S removal due to operation at low pH; increasing pH by NaOH addition was not feasible because of bicarbonate precipitation. [11] Switching to KOH resolved the problem, but it can be 1.5 to 3 times as expensive as NaOH on a mole per mole basis.
Sulfur quality. In certain solvents, elemental sulfur behaves well; however, water is not one of these solvents. Sulfur is extremely insoluble in water ([approx][10.sup.-7] gram moles/liter). As a result, the elemental sulfur tends to nucleate uncontrollably throughout the aqueous liquid. Nucleation results in very tiny particles, often in the range of 1-20 microns.
The sulfur formed in most aqueous systems is also hydrophobic; that is, the tiny sulfur particles float on the liquid surface because water will not wet them. Surfactants are added to wet them and allow them to adhere to each other, forming aggregate particles large enough to sink and be removed. In other cases, where bacteria are present, the surfactant may not have to be added because the bacteria produce surfactants that aid in wetting the sulfur particles. In general, surfactants on the sulfur tend to attract other contaminants such as hydrocarbons, which can be incorporated into the clump of tiny particles. This type of material is difficult to separate cleanly from the process solution, typically forming a paste in filters and centrifuges. One technology developer candidly characterized the 50-70 wt% sulfur cake product as “viscous sludge.” [12]
Fig. 3 compares sulfur formed in an aqueous system to sulfur crystallized in the new, nonaqueous process (discussed later). The sulfur from the aqueous system consists mostly of [approx] 5-micron particles that are stuck together in larger accumulations. Contaminants (hydrocarbons, salts, etc.) can be contained within the surfactant layer that coats these particles and in the space between particles. The sulfur from the nonaqueous process consists primarily of large crystals of pure elemental sulfur.
Plugging. The presence of solids creates a potential for plugging. The surfactants that make the sulfur particles adhere can also make the sulfur sticky, resulting in equipment plugging on a frequent basis, as has been documented numerous times. [7,8,9] Very high liquid flows seem to postpone incidents of plugging; however, high liquid rates result in high pumping costs for even moderate-pressure applications. Designers of low-pressure processes that normally have very high liquid flowrates may tend to propose reducing the liquid flowrate as much as possible for high-pressure applications to reduce pumping costs. However, incidents of plugging are more likely.
Foaming. Both surfactants and tiny particles cause foaming. [13] There is evidence that foaming in aqueous systems may be most severe in high-pressure applications. [7] The particular combination of surfactants and tiny particles found in aqueous sulfur recovery systems is one reason for this increased foaming tendency. Tiny bicarbonate particles formed in the absorbers of sodium-based aqueous systems at high [CO.sub.2] partial pressures could also stabilize foams.
Materials. Aqueous systems are naturally more corrosive to metals than many nonaqueous systems. All aqueous solutions are electrically conductive, which can greatly affect corrosion rates. Some contain chelates that dissolve iron. For these reasons, all aqueous sulfur recovery processes require corrosion-resistant materials such as stainless steel (especially in the absorber area). Lined carbon steel, plastic and concrete may be used for some of the atmospheric process operations on certain smaller-scale applications.
Aqueous-iron liquid redox processes. As with all the aqueous sulfur recovery systems, aqueous-iron systems were first developed and applied on low-pressure gas streams. With an aqueous-iron process, gas containing [H.sub.2]S contacts scrubbing liquor in a liquid-full absorber. The scrubbing liquor is chelated iron in an aqueous slurry of elemental sulfur particles. The [H.sub.2]S is converted to additional elemental sulfur particles in the absorber, and sweet gas exits the top. After the rich liquor slurry is regenerated, it is returned to the absorber, and a portion is withdrawn to separate the sulfur particles. Several variations of the process flow scheme exist.
Aqueous-iron processes are generally capable of doing a good job with the first of the desirable features for high-pressure natural gas treatment: [H.sub.2]S removal to less than 4 ppm. Fast, irreversible chemical reactions convert [H.sub.2]S into elemental sulfur within the absorber. The fast, irreversible nature of the reactions enhances [H.sub.2]S removal capabilities by eliminating [H.sub.2]S backpressure within the slurry.
However, aqueous-iron redox systems lack any of the remaining desirable characteristics. [CO.sub.2] affects both operability and chemical costs. Most aqueous-iron systems use dilute chelate solutions; under high [CO.sub.2] partial pressures, these dilute systems need more expensive KOH (rather than NaOH) for pH control to avoid formation of bicarbonate solids.
Liquid circulation rates are high compared to most other systems, as the accompanying table shows. Tiny sulfur particles, which are present throughout the system, cause chronic problems with positive displacement pumps and are the root cause of severe plugging and foaming problems. The chelate solution requires stainless steel equipment, which can corrode severely in areas where sulfur solids deposit and provide crevice-like conditions. Finally, aqueous-iron processes have inherent chelate degradation problems as a result of the formation of aggressive species hydroxyl radicals) in the regeneration [step.sup.14 ,5]
Aqueous-iron processes were shown to be unacceptable for direct treatment of high-pressure sour natural gas during tests sponsored by GRI (now GTI, Gas Technology Institute) where two different commercial aqueous-iron processes sought to treat a 1,000-psig sour natural gas [stream.sup.7,8] The most severe inherent problem with aqueous-iron systems is the immediate formation of sulfur solids in the absorber and the subsequent presence of sulfur solids circulating throughout such systems. During the GRI tests, both processes failed to operate reliably for more than one to two weeks at a time, even with technology licensors’ staff on-site directing the plant Operators.
Aqueous caustic processes with bacterial regeneration. As with the aqueous-iron systems, this type of process was developed and applied primarily on low-pressure streams, with original applications, and most of the 10 to 20 subsequent applications, on atmospheric wastewater offgas (biogas), primarily in Europe. Other than the wastewater applications, the literature commonly mentions only one other plant that went into commercial operation. This plant operated on a low-pressure contaminated air steam in a chemical plant for about two years, between 1997 and 1999. [16] A second non-wastewater, low-pressure application is planned for startup in 2002. [17] Recently, this type of process has been proposed for processing high-pressure natural gas.
A caustic/bacteria process does have some of the desirable features for high-pressure operation. It may be possible to remove [H.sub.2]S to pipeline specs for cases with low [CO.sub.2] partial pressures. Another positive is the reported low instance of plugging with the sulfur excreted by the bacteria. Otherwise, this type of process does not have any of the other desirable characteristics. Since it operates between pH 8 and 9, sodium bicarbonate precipitation could be a problem at high [CO.sub.2] partial pressures.
Amine/aqueous-iron redox combinations. Fig. 4 shows a simplified flow diagram for an amine/aqueous-iron liquid redox combination for treating high-pressure natural gas. As with the amine/Claus/TGT combination, the amine unit absorber is the only high-pressure vessel. Essentially all the [H.sub.2]S and some or all of the [CO.sub.2] are absorbed into an aqueous alkanolamine solution in the absorber. These acid gases exit the amine system in the low-pressure stripper column and are then fed to the aqueous-iron redox unit, which converts essentially all of the [H.sub.2]S to elemental sulfur particles.
This combination has the advantage of using well-proven amine technology for the high-pressure absorber. The aqueous-iron redox portion of the plant operates near atmospheric pressure, so plugging problems will be limited to the low-pressure side of the plant. As a result, this combination has most of the desirable features listed previously, but involves additional cost. There is still some effect of higher [CO.sub.2] concentrations on amine unit design, on the buffer system used for the aqueous-iron system (to avoid bicarbonate precipitation), and on the amount of caustic needed to control pH in the presence of high [CO.sub.2] concentrations.
The nonaqueous CrystaSulf process. Fig. 5 shows a simplified flow diagram for the new process as applied to high-pressure, sour natural gas. Although it can be applied to low-pressure situations, it was developed specifically to treat high-pressure sour natural gas. The process uses a nonaqueous solution with a high solubility for elemental sulfur. Because the elemental sulfur stays dissolved in the solution, there are no solids in the liquid circulated to the absorber. By design, CrystaSulf avoids the problems that make aqueous sulfur recovery systems unsuitable for direct treatment of high-pressure gas.
With this process, [H.sub.2]S is removed from the sour gas in a conventional tray absorber. The [H.sub.2]S reacts with dissolved sulfur dioxide ([SO.sub.2]) to produce dissolved elemental sulfur. There are no solids in the absorber, flash tank, or solution lines, which means there is no chance of plugging. Rich solution from the absorber passes to a flash step. Depending on the processing conditions, it may at times be economical to compress the flash gas and return it to the inlet stream.
After the flash step, the solution flows to a crystallizer, where temperature is lowered and solid elemental sulfur crystals form. The crystallizer/filter area is the only area where sulfur solids exist within the process. The crystallizer overflows to a surge tank. A heater in the surge tank ensures that all elemental sulfur is in a dissolved state. A conventional positive displacement pump transfers the solution back to the absorber.
For high-pressure applications, [SO.sub.2] can be added two different ways. Pure [SO.sub.2] can be purchased and metered into the lean solution line; this option is economical where [SO.sub.2] is readily available for purchase and sulfur throughputs are small. Alternatively, a portion of the product sulfur can be burned, and the resulting [SO.sub.2] can be absorbed into the CrystaSulf solution via a separate small [SO.sub.2] absorber.
In this process, [SO.sub.2] binds chemically with species in the nonaqueous solution. The bond is strong, and there is generally no detectable concentration of [SO.sub.2] in the gas phase anywhere within the system, including the sweetened gas. A large quantity of [SO.sub.2] can exist within the solution, and this background concentration creates a buffering effect.
The process has all seven desirable features needed for high-pressure natural gas treatment. It has the ability to remove [H.sub.2]S to below pipeline gas specs. High [CO.sub.2] partial pressures have no effect on the process, and [CO.sub.2] is not removed. A pilot unit has operated for over 4,000 hr on an 80% [CO.sub.2], 300 psig application (16 atmospheres of [CO.sub.2] partial pressure). Circulation rates are low, similar to those used for aqueous alkanolamine systems (see table). Since there are no solids in the solution, efficient positive displacement pumps can be used. The solution does not foam because there are no surfactants or particles present.
Since the solution dissolves elemental sulfur, there are no solids present (except in the crystallizer/filter section) and hence no plugging. The solution is also noncorrosive. Extensive corrosion testing has been performed at the pilot unit shown in Fig. 6, located at a commercial site in West Texas. Additional process details are available. [18] To date, commitments have been received for two commercial CrystaSulf units, and initial design work has begun on the first one.
COST COMPARISONS
A cost comparison was performed for three options of treating high-pressure sour natural gas: 1) MDEA/ Claus/TGT, 2) amine/aqueous-iron, and 3) CrystaSulf (CS). The base case for raw gas characteristics was treatment of 40 MM Scfd of gas at 1,000 psig, 0.2 mol % [H.sub.2]S and 3 mol % [CO.sub.2] Methyldiethanolamine (MDEA) was used as the amine in all amine cases. The aqueous-iron unit was assumed to be of a simple, auto-circulation style design. The Claus unit costs assumed a straight-through unit with two converter beds. The tail-gas treater was assumed to be a single-bed direct oxidation unit. CS unit costs were based on purchasing pure [SO.sub.2] rather than burning sulfur. Modular construction was assumed for all units. Total treating cost was taken to be the sum of the operating costs, plus one-fifth of the capital cost (20% capital recovery factor).
As can be determined from Fig. 7, CS has a significant cost advantage over both MDEA/aqueous-iron and MDEA/Claus/TGT options for treatment of gas with 3 mol % [Cosub.2]. Capital cost for the amine/aqueous-iron and MDEA/Claus/TGT options were 175% and 260%, respectively, of CS’s capital costs. Operating costs for the amine/aqueous-iron case were 188% of CS operating costs, resulting in a total annual cost greater than 180% of the CS case. While operating cost for the MDEA/Claus/TGT case is only about 110% of CS’s, the overall annual cost for MDEA/Claus/TGT is still about 180% that of CS, because of the 20% capital recovery factor. Details and background information on the cost study and other comparative data are available from the authors. 18
CONCLUSIONS
Sulfur recovery from high-pressure sour natural gas streams with medium amounts of sulfur has historically been a challenge for the industry. Treatment with traditional amine/ Claus/TGT approaches is very capital intensive. Direct treatment with aqueous-iron liquid redox processes was once proposed as a more economical processing option; but high-pressure operation of these systems proved infeasible due to severe, inherent operating problems. Other proposed systems, e.g., caustic systems with bacterial regeneration, were developed for low-pressure applications and may have difficulty operating at high [Co.sub.2] partial pressures because of sodium bicarbonate precipitation.
CrystaSulf was developed specifically for treating high-pressure natural gas streams with medium amounts of sulfur. It is not affected by high [CO sub.2] partial pressures and solves the problems of the aqueous systems by eliminating the presence of sulfur solids in the highpressure equipment. The process is being marketed by Crysta Tech, Inc., Austin, Texas. Initial design work has begun on the first of two commercial units. Additional information is available on the Crysta Tech website, www.crystatech.com.
Kenneth E. McIntush, P.E., is VP of engineering and a founder of CrystaTech, Inc., a company formed in 2000 to develop/provide process technology. He led engineering aspects of the CystaSulf development effort during his last five years at Radian International, LLC (now URS Corp.), focusingon [H.sub.2]S removal and sulfur recovery for over 10 years. This work involved engineering, design and construction of largescale pilot units to test several sulfur recovery technologies for high-pressure gas treating. He plans and implements Crysta Tech’s engineering. Mr. Mclntush holds a BS in chemical engineering from Texas A&M University, College Station. He is a registered professional engineer in Texas, and has authored several papers on sulfur recovery.
Dennis A. Dalrymple, P.E., is president and CEO of Crysta Tech, Inc. While at Radian International, he helped establish the firm’s sulfur recovery business 15 years ago and led its evolution from technology evaluation to development/commercialization. He served as department head of process engineering, then gas processing business manager. Prior to Radian, he worked in engineering/management positions with E.I. duPont and Conoco. He was a process engineer at duPont’s textile fibers plant in Waynesboro, Virginia. With Conoco, he held process/project engineering positions at the Houston refining headquarters and Westlake refnery, Lake Charles, Louisiana, where he led installation of a fluid catalytic cracking unit. Mr. Dalrymple is a registered professional engineer in Texas, and holds a BS in chemical engineering from the University of Texas, Austin, and an MBA from Southwest Texas State University.
Curtis 0. Rueter, RE., serves as VP of marketing and natural gas sales of Crysta Tech, Inc. Since the firm’s founding he has led the marketing/sales effort. He has made numerous presentations at national/regional gas processing meetings. Prior to the founding of Crysta Tech, he provided engineering support to Radian International’s sulfur recovery research. And he led commercialization of a process for controlling glycol dehydrator emissions, including distribution of several software packages. At Radian, he served for four years as site manager for a branch office. Mr. Rueter is a registered engineer in Texas and holds a BS in chemical engineering from Texas A&M University, College Station.
LITERATURE CITED
(1.) Dalrymple, D. A., F. D. Skinner and N. P. Meserole, Investigation of U.S. natural gas reserve demographics and gas treatment processes, Topical Report, GRI-91/0019, Gas Research Institute. Chicago, Illinois. Section 3.0. pp. 3-1-3-13, 1991.
(2.) Hugman, R. H., P S. Springer and E. H. Vidas, Chemical composition of discovered and undiscovered natural gas in the United States: 1993 update, Topical Report, GRI-93/0456.l-3. Gas Research Institute, Chicago, Illinois, 1993.
(3.) Cornot-Gandolphe, S., “Changes in world gas reserves and resources,” Energy Exploration Exploitation, 13(1): 3-16, 1995.
(4.) Quainlan M., “Evaluation of selected emerging sulfur recovery technologies,” GRI Gas Tips, 3(1),26-35 (Winter 1996/1997).
(5.) Meyer, H.,D. Dalrymple and K. Mcintush, “Opportunities for improvements in offshore gas processing,” presented at the 75th Annual GPA Convention, Denvet, Colorado, 1996.
(6.) Fisher, K. S., J. E. Lundeen and D. Leppin, Fundamentals of [H.sub.2.]S scavenging for treatment of natural gas,” presented at The Ninth GRI Sulfur Recover Conference, October 24-27 1999, San Antonia, Texas.
(7.) McIntush, K. E. and B. J. Petrinee, GRI resting of SulFerox for the direct treatment of higit-pressttre natural gas at NGPL’s Kermit, Texas site, Final Report, GRI-94/0432 Gas Research Institute, Chicago, Illinois, 1995.
(8.) Holloway, C. S., GRI testing of ARI-LO-CAT H for the direct treatment of high-pressure natural gas at NGFL’s Kermit. Texas site. Final Report, GRI-96/0007. Gas Research Institute, Chicago, Illinois, 1996.
(9.) Dalrymple, D. A. and J. K. Wesels, “Liquid redox users panel handouts,” Proceedings, 1995 GRI Sulfur Recovery Conference, Topical Report, GRI-95/0441, Gas Research Institute, Chicago, Illinois, pp. 187-271.
(10.) Anon ACAP Aquenous chemical and physical properties. Version 2.1, Copyright 1987, P.S. Lowell and Co., Inc., Austin, Texas.
(11.) Reicher, M., B. Niemiec and T. Katnna, Modifications solve dmulfurization uuit start-up problems,” Oil and Gas Journal, 91(11). pp. 55-59. March 13, 2080.
(12.) Anon., Paques Journal newsletter, Autumn 1994, Faques BV, Balk, the Netherlands, p.7.
(13.) von Phul, S.A., “Sweetening process foaming and abasement,” Laurance Reid Gas Conditioning 2001. Conference Proceedings, Norman, Oklahoma, Feb. 25-28, 2001, pp 251-280.
(14.) DeBerry, D. W. and D. M. Seeger, “Mechanisms of chelant degradation in iron based liquid redax processes.” 1992 Liquid Redox Sulfur Recovery Conference, Austin, Texas. Oct. 1992.
(15.) DeBerry D., “Chemical evolution of liquid redox processes.” Environmental Progress, 16(3), 1997.
(16.) Janssen, A.J.H., B. J. Arena and S. Kijlstra, New developments of the THIOPAQ process for the removal of [H.sub.2]S from gaseous streams,” preprints of Sulphur 2000 Conference, Oct. 29-Nov. 1 2000, San Francisco, California, pp. 179-187.
(17.) Kijlstra, S., A. Janssen and B. Arena, Biological process for [H.sub.2]S removal from (high-pressure) gas: The Shell-Paques/THIOPAQ gas desulfurization process,” Lanrance Reid Gas Conditioning Conference 2001, Conference Proceedings. pp. 169-182.
(18.) McIntush, K. B., D. M, Seeger. C. 0. Roeter and O.W DeBerry, ‘Comparison of technolgies for removing sulfur from high-pressure sour natural gas with sulfur throughputs between 0.1 and 30 long tons/day,” presented at the 80th Annual GPA Couvention, San Antonio, Texas, March 2001.
Comparison of circulation rate to [H.sub.2]S absorber [(0.5 vol.%
[H.sub.2]S, 2.0 vol.% [Co.sub.2]).sup.18]
System gpm/LTPD
Alkanolamine (MDEA) 30-60
Aqueous-iron 200-1,200
Caustic with bacterial regeneration 300-3,000
CrystaSulf 20-50
[Graph omitted]
[Graph omitted]
COPYRIGHT 2001 Gulf Publishing Co.
COPYRIGHT 2001 Gale Group
